Professional Documents
Culture Documents
Badmintostatsanalysis
Badmintostatsanalysis
Department of Mechanical and Electronical Engineering, China Agriculture University, Beijing 100083, China
Abstract: The use of computer vision technology to collect and analyze statistics during badminton matches
or training sessions can be expected to provide valuable information to help coaches to determine which
tactics should be used by a player in a given game or to improve the player’s tactical training. A method
based on 2-D seriate images by which statistical data of a badminton match can be obtained is presented.
Image capture and analysis were performed synchronously using a multithreading technique. The regions of
movement in the images were detected using a temporal difference method, and the trajectories of the
movement regions were analyzed using seriate images. The shuttlecock trajectory was extracted from all
detected trajectories using various characteristic parameters. The stroke type was determined by comparing
the shuttlecock trajectory data with a set of stroke definition data. The algorithm was tested at a training cen-
ter, and the results were compared with baseline data obtained by expert visual inspection using four video
samples, which included approximately 10 000 frames. The shuttlecock trajectory and stroke type were de-
tected correctly in almost 100% of the analyzed video sequences. The average speed of the automated
analysis was approximately 40 frames/s, indicating that the method can be used for real-time analysis dur-
ing a badminton match. The system is convenient for use by a sports coach.
Key words: machine vision; 2-D measurement; badminton match; trajectory tracking; stroke type statistic
players during a soccer game through the use of multi- (5) Calculation of the statistics for the stroke type by
ple cameras[11]. Ye et al. have proposed a method for comparing the characteristics of the detected shuttle-
athlete identification by combining the segmentation, cock trajectory and the definition of the stroke type.
tracking, and recognition procedures into a coarse-to-
fine scheme for jersey number (digital characters on 1.1 Equipment
the sport shirt) detection[12]. Wang et al. proposed an
A Basler A601f series progressive scan CMOS (com-
automatic technique for extracting color models of the
plementary metal-oxide-semiconductor transistor)
playing surface and the team uniforms that can be used
camera was used to acquire sample images. The cam-
in higher-level processes such as tracking and recogni-
era was designed for industrial applications. The pri-
tion[13]. Qi et al. presented a framework for tracking
mary features of the camera are as follows. The camera
sports players in videos recorded by static domestic
matches the 1394 TA (transmission adapter) digital
cameras[14]. Zaveri et al. proposed an algorithm for de-
camera specification. The maximal frame rate at full
tection and tracking of small and fast moving objects,
resolution is 60 frames/s. The asynchronous full frame
such as a pingpong or a cricket ball, in sports video se-
is shuttered electronically. The camera has square pix-
quences[15].
In badminton matches and in training, in order to se- els, and the frame size is 659×493 pixels. The camera
lect the tactics required to deal with different oppo- was used to capture monochrome images.
nents, the coach must have a good statistical know- Capturing and processing of each image were per-
ledge of the tactics of the player. The ultimate goal of formed using a notebook personal computer with a
this study was to develop a system capable of real-time Pentium 2.4-GHz CPU. The algorithm was developed
statistical analysis of data collected either during train- based on the platform of a 2-D moving image meas-
ing or match-play, to provide information to assist urement system, which was provided by Beijing Mod-
training and to select match tactics. ern Fubo Technology Ltd., implemented using Micro-
The use of 3-D information should be able to pro- soft Visual C++ 6.0.
vide accurate statistical data for analysis of a badmin-
1.2 Sampling a moving shuttlecock
ton match. However, obtaining 3-D information re-
quires the use of more than two cameras, and real-time 1.2.1 Preparation for the sampling program
analysis requires high-capacity processors, the size and The application for synchronous capture, analysis, and
volume of which are generally too large for a practical data storage uses Windows multithreading technology.
implementation. Therefore, 3-D detection is not suit- When the application is started, the captured thread is
able for badminton coaches, as the equipment must be started and captured images are displayed on the
readily, transportable to remote match locations. In the screen. Then, when analysis and/or saving has begun,
present study, therefore, we develop a 2-D measure- thread analysis and/or saving is started, and image
ment and analysis system, which requires only one analysis and/or saving is carried out synchronously.
camera and a laptop personal computer. 1.2.2 Establishing the camera and demarcating
the badminton court
1 Method
In the present study we attempt to develop a real-time
The system can be broken into a number of steps as measurement algorithm for a moving shuttlecock in an
follows. actual match. Therefore, the image samples used in the
(1) Capture of 2-D images of the badminton court present study should simulate scenes of an actual
that are suitable for use in measuring the shuttlecock match as closely as possible. In order to capture the en-
trajectory using one camera. tire scene of the badminton court, the camera was
(2) Extraction of moving objects from 2-D seriate mounted 5 m from the corner of the court at a height of
images. 4 m from the ground and at an angle of approximately
(3) Classification of moving objects and connection 45q from the vertical.
of their tracks. Before the image analysis, eight characteristic coor-
(4) Extraction of the shuttlecock trajectory from dinates of the badminton court were selected
entire tracks. manually, as shown in Fig. 1. The representational
596 Tsinghua Science and Technology, October 2007, 12(5): 594-601
points comprise the two top ends of the badminton net The outline of each white pixel block is measured on
(point 1,3), the two feet of the net frame (point 2,4), the binary image. According to the coordinates of the
and the four corners of the court near the net (point 5-8). outline, the barycenter of each white pixel block is cal-
culated as
1 n 1 1 n 1
X ¦
ni 0
xi , Y ¦ yi
ni 0
(3)
trajectory up to this point is analyzed in order to de- position of the trajectory start point, the position of the
termine whether the longest trajectory is the shuttle- trajectory end point, the direction and the distance of
cock trajectory. the line from the trajectory start point to the end point,
the trajectory camber, the average distance between
1.4 Detecting the shuttlecock track
points on the trajectory, the location of the highest
Tmax denotes the longest trajectory among all of the point on the trajectory, the distance between the trajec-
tory start point and the end point to the players, the dis-
trajectories, and O1 , O2 ,..., Om denote all of the proper-
tances from the start point and the end point to the
ties that are used to evaluate whether a trajectory is the
badminton net, the direction number of the trajectory,
shuttlecock trajectory. These properties include the tra-
the coordinates of the badminton net, the coordinates
jectory length, the trajectory camber, and the trajectory
of the court contour, and the locations of the trajectory
direction number. The reliability parameter \ (Tmax )
start and end points in the badminton court.
of the shuttlecock trajectory is defined as Flow chart of the detection system is shown in Fig. 2.
m
\ (Tmax ) ¦\
i
i (Oi ) (4)
The detected trajectory data were compared with a video samples, which included a total of approximately
set of stroke type definitions to determine the stroke 10 000 scenes. The results are compared in Table 1.
type. The results obtained by the algorithm were com- Video samples 1, 2, 3, and 4 had 394, 1204, 3108, and
pared with those obtained from visual inspection of the 6020 scenes, respectively.
Table 1 Measurement results by the algorithm (A) and visual inspection (V)
Lob Raise Smash Backstop Punch Whip
Video sample
A V A V A V A V A V A V
1 2 2 2 2 1 1 1 1
2 2 2 9 9 3 3 1 1 6 6 1 1
3 9 9 9 9 2 2 3 3 8 8 20 20
4 6 6 28 29 16 16 6 6 13 13 4 4
As shown in Table 1, the detection rate of the shut- track is greatly disturbed; (2) improvement of the algo-
tlecock track was close to 100%. In only one case, a rithm to allow a more detailed definition of the stroke
raise stroke, the shuttlecock track was not recorded. In type; and (3) improvement of the algorithm to enable
this case, the shuttlecock trajectory was in fact deter- the score of the badminton match to be judged.
mined but ignored because the shuttlecock trajectory
Acknowledgements
was not the longest trajectory among the group of tra-
jectories. This study supported by Beijing Modern Fubo Technology Ltd.
The test results indicate that the algorithm can be We thank Dr. Sun Ming, at China Agriculture University, Ms.
used to obtain statistical data of player tactics during a Wang Kaizhen and Mr. Zhou Zhihui, at Beijing Sports Univer-
badminton match. sity, and Mr. Chen Yao, at the Pingpong and Badminton Center
of China National Sports Ministry, for their help in this study.
3 Conclusions
References
This paper described a method to obtain statistical data
of a badminton match based on 2-D seriate images. [1] Pingali G, Opalach A, Jean Y. Ball tracking and virtual re-
Moving portions of the images, including the shuttle- plays for innovative tennis broadcasts. In: Proceedings
cock, rackets and players, are detected and labeled as 15th International Conference on Pattern Recognition.
white pixel blocks based on an image difference Washington, DC, USA: IEEE Compter Society Press,
method. The outline of each white pixel block is meas- 2000: 152-156.
ured in the binary image, and the barycenter is calcu- [2] Pingali G, Jean Y, Carlbom I. Real-time tracking for en-
lated from this outline data. Each barycenter point is hanced sports broadcasts. In: Proceedings of CVPR98.
connected to a moving track according to several data. Washington, DC, USA: IEEE Compter Society Press,
The shuttlecock trajectory is extracted from the trajec- 1998: 260-265.
tory data. After the shuttlecock trajectory is determined, [3] Yu Xinguo, Xu Changsheng, Leong Hon Wai, et al. Tra-
the stroke type is determined by comparing the trajec- jectory-based ball detection and tracking with applications
tory data to the definition data for various stroke types. to semantic analysis of broadcast soccer video. In: Pro-
This process is continued until the end of each badmin- ceedings of the 11th ACM International Conference on
ton match. The output is a set of statistical data of the Multimedia. New York, NY, USA: ACM Press, 2003:
stroke type for each player. 11-20.
The algorithm can be used not only for the training [4] Yu X, Tian Q, Wan K W. A novel ball detection frame-
and sports match, but also for other automated analysis, work for real soccer video. In: Proceedings of ICME.
such as the track of the insects. In the future, we hope Washington, DC, USA: IEEE Computer Society Press,
to accomplish three objectives: (1) improvement of the 2003: 265-268.
algorithm to identify the objective track even when the [5] Yu X, Xu C, Tian Q, et al. A ball tracking framework for
CHEN Bingqi (чρ୫) et alġA Statistical Method for Analysis of Technical Data of … 601
broadcast soccer video. In: Proceedings of ICME. Wash- [11] Figueroa P, Leite N, Barros R M L, et al. Tracking soccer
ington, DC, USA: IEEE Computer Society Press, 2003: players aiming their kinematical motion analysis. Com-
273-276. puter Vision and Image Understanding, 2006, 101(2):
[6] Yu Xinguo, Sim Chern-Horng, Wang Jenny Ran, et al. A 122-135.
trajectory-based ball detection and tracking algorithm in [12] Ye Qixiang, Huang Qingming, Jiang Shuqiang, et al. Jer-
broadcast tennis video. In: Proceedings of ICIP. New York, sey number detection in sports video for athlete identifica-
NY, USA: ACM Press, 2004: 1049-1052. tion. In: Proceedings of SPIEüThe International Society
[7] Ekin A, Tekalp A M, Mehrotra R. Automatic soccer video for Optical Engineering, Visual Communications and Im-
analysis and summarization. IEEE Transactions on Image age Processing. 2005: 1599-1606.
Processing, 2003, 7(12): 796-807. [13] Wang Lei, Zeng Boyi, Lin S, et al. Automatic extraction of
[8] Utsumiy O, Miurayy K, Idez I, et al. An object detection semantic colors in sports video. In: IEEE International
method for describing soccer games from video. In: IEEE Conference on Acoustics, Speech and Signal Processing.
International Conference on Image Processing. 2002: Canada, 2004: 617-620.
45-48. [14] Qi Fei, Luo Yupin, Hu Dongcheng. Visual tracking of
[9] Perš J, Kovacic S. Computer vision system for tracking players through occlusions in low resolution. In: 6th
players in sports games. In: Proceedings of Int’l Work- IASTED International Conference on Signal and Image
Shop Image Signal Processing Analysis. Pula, Croa-tia, Processing. Singapore, 2004: 375-380.
2000: 177-182. [15] Zaveri M A, Merchant S N, Desai U B. Small and fast
[10] Perš J, Kovacic S. Tracking people in sport: Making use of moving object detection and tracking in sports video se-
partially controlled environment. In: Proceedings of the quences. In: IEEE International Conference on Multimedia
9th International Conference on Computer Analysis of Im- and Expo. Taipei, China, 2004: 1539-1542.
ages and Patterns. London, UK: Springer-Verlag, 2001:
374-382.